Lesson 6 of 6

The Only Metrics That Matter at Idea Stage

Forget MRR and ARPU for now. Early on you track signal: commitment, retention, and pull.

7 min read

Open any startup metrics guide and you drown in acronyms: MRR, ARPU, LTV, CAC, churn, the rule of 40. They matter enormously later. At the idea stage, most of them are a distraction, because you do not have enough customers for any of them to mean anything, and tracking them is often a way to feel busy while avoiding the one scary question: does anyone actually want this. Early on you measure signal, not revenue.

Vanity metrics feel good and tell you nothing

A vanity metric is any number that goes up and to the right without telling you whether you have a business. Page views, social followers, signups, likes. They are seductive because they are easy to grow and they feel like progress, and they are dangerous because you can have a lot of all of them and still be building something nobody will pay for. If a number can rise while your actual business goes nowhere, it is vanity.

The cure is to ask, for any metric you track, what decision it would change. If the answer is none, it just feels nice, drop it. The numbers worth your attention at this stage are the ones tied to real behavior and real commitment: did they come back, did they do the core action again, did they pay. Those can ruin your week, which is exactly why they are worth watching.

Track signal: commitment, retention, and pull

The three signals that matter before you have a product are simple. First, commitment: are people giving up something real, money, a pre-order, serious time, not just a click. Second, retention: of the few users you have, do they come back and do the core thing again without you nagging them, because a leaky bucket does not get better at scale, it gets worse. Third, pull: are people asking for more, referring others, frustrated when it is not ready. Pull is the feeling that you are being dragged forward instead of pushing a rock uphill.

These are not dashboard metrics, they are mostly things you notice by being close to a handful of users. That closeness is a feature of the idea stage, not a bug. You will never again have so few customers that you can feel each one. Use it. The qualitative sense of whether people pull or drift is more predictive of whether you have something than any ratio you could compute from twelve data points.

Commitment: pre-orders, payments, signed pilots. Not clicks.
Retention: do the few users you have come back and repeat the core action?
Pull: do people ask for more, refer others, complain when it is down?
The test of any metric: what decision would it change? If none, ignore it.

Know the money metrics exist, then move on

This is not permission to stay innumerate. You should understand what the big metrics mean, because the moment you have paying customers they become the whole game, and a couple of them are worth a back-of-envelope check even now. It is worth knowing roughly what a customer might be worth over their lifetime against what it might cost to acquire one, because if those numbers can never work, no amount of validation saves the idea. A quick sanity check beats a precise spreadsheet built on imaginary inputs.

The same goes for pricing models like freemium, where most users pay nothing and a few convert, and revenue-per-user measures like ARPU. You do not need to optimize them at the idea stage, you need to know they exist and roughly where your idea would land, so you are not blindsided later. Once you have real paying customers and real retention, graduate to the full picture. Until then, signal first, spreadsheets second.

Worked example

A leaky bucket does not improve at scale

Imagine two idea-stage founders. The first is thrilled: ten thousand signups in a month from a viral post. The second is quieter: forty users, but thirty of them come back every week and five have already paid for early access. On a vanity dashboard the first founder is winning by two hundred to one. In reality the second has the business, and the first has a traffic spike.

The reason is retention. If signups pour into a bucket that leaks, scaling the traffic just leaks faster, and you spend money pouring water into a hole. The second founder has proven the bucket holds: people come back, some pay, the value sticks. That is the thing you scale. The first founder has proven only that a headline worked once. Fix retention at forty users, then go find ten thousand.

Learn by doing

Paste these into ChatGPT or Claude and run them against your own idea. The model will answer happily. Olune goes further and checks the answer against real Reddit threads, competitor maps, and keyword volume.

Tip: grab the downloadable playbook to run every play with fill-in worksheets →

Prompt 1 · Strip your idea down to the one signal that matters now.

I am at the idea or early-prototype stage with [describe your idea and where it is]. I keep getting distracted by vanity metrics. Act as a no-nonsense coach. Tell me the single most important signal I should be tracking right now given my stage, why it beats the others, and the simplest way to measure it without a fancy dashboard.

What a good output looks like

It might say: you have twenty users. Forget signups and traffic. Track week-over-week return on the core action, measured by hand in a spreadsheet. If they do not come back, nothing else matters, and you can see it with twenty rows.

Prompt 2 · Sanity-check whether the economics could ever work.

Here is my idea and a rough price: [describe it and what you might charge]. Do a back-of-the-envelope check on whether the unit economics could plausibly work. Estimate a rough lifetime value and a rough cost to acquire a customer for this kind of product, state your assumptions, and tell me if there is an obvious reason the math could never add up.

What a good output looks like

You get a rough lifetime value from price times a plausible retention, a ballpark acquisition cost for the channel, and a verdict like at 9 a month with likely churn, lifetime value is around 60, so paid ads at 40 a customer are tight but not impossible, while enterprise sales would never pay off. Directional, not precise.

Key terms in this lesson

Takeaways

At idea stage you measure signal, not revenue. Most acronyms are a distraction with twelve data points.
Vanity metrics rise while the business goes nowhere. If a number changes no decision, drop it.
Watch three things: commitment, retention, and pull. They can ruin your week, which is why they are honest.
Know LTV, CAC, freemium, and ARPU exist and roughly where you land. Optimize them once you have real customers.

Now run your own idea through it.

You have the method. Olune does the legwork: an honest build-or-kill verdict on live Reddit signals, competitor maps, and keyword volume, in about eight minutes. Free to start.

Test your idea →

Common questions

Should I really ignore revenue at the idea stage?

Not ignore, reframe. Early revenue matters enormously as a signal, because someone paying is the strongest validation there is. What you ignore is optimizing revenue metrics like ARPU or growth rates when you have a handful of customers, because the numbers are too small and noisy to optimize. Treat early money as proof, not as a dashboard to tune.

What is the difference between a vanity metric and a real one?

Ask what decision the number would change. A real metric, if it moved, would make you do something different: keep going, pivot, fix retention, change the price. A vanity metric just feels good and changes nothing. Signups can be either, depending on whether you act on them. Tie every number to a decision or stop tracking it.

When do LTV, CAC, and the rest start to matter?

Once you have real paying customers and enough retention data to trust the inputs, usually after you have found a repeatable way to get and keep customers. At that point unit economics become the core of the business and deserve real attention. Before then, a rough sanity check is enough.